Against 2d6

The players from my Classic Traveller one-shot last week started asking whether there would be more adventures for those characters.  I felt a certain unease at the prospect, and decided to examine it.

I think I really dislike 2d6 as a core resolution mechanic, and "fighting with" the dice last session reminded me why.

The central tendency is very strong.  You get many results right in the middle of the range, and small modifiers can change the probability of success dramatically when the target number is close to the center of the range.  The total range is also quite small (2-12), so it really doesn't take that many small modifiers stacking up to make success or failure all but certain.  This is the same problem that ACKS 1e ran into with stacking up modifiers to reaction rolls, except that every piece of Traveller is susceptible to this failure-mode.  This was one nice thing about Thousand Suns' move to 2d12 - it becomes much harder for modifiers to completely dominate randomness (apparently I wrote a draft post of a review of Thousand Suns back in 2015 but never published it - oops).

In theory, many Classic Traveller skill rolls have modifiers from attributes and skill level laid out in the text, under the description of the skills.  However, in a time-constrained after-work session, digging out the skill descriptions for every roll is prohibitively time-expensive.  Even in a longer session, it stalls the tempo.  So I am inclined to improvise target numbers.  In an OSR game without a skill system, I would pick probabilities which sound reasonable within the game-world and negotiate with the players.  "That sounds pretty challenging, maybe a 20% chance of success"  "I'm very strong though."  "OK, 35% - give me a 14+ on a d20".  Translating intuitive probabilities of success into 2d6 target numbers is a huge pain, and the central tendency means that there is very little room for fine-grained distinctions between target numbers in the middle of the range.  You go from 72% chance of success on 6+ to 58% chance of success on 7+ to 42% on 8+.  You can only draw fine-grained distinctions out at the edges of the range (11+ vs 12+, for example; 8% vs 3%).  Even there, the granularity of target numbers is no finer than a d20's.

I think the only real virtues of 2d6 systems are that the player-facing math is easy (almost no significant two-digit operations - if your roll plus skill get into two digits, success is a foregone conclusion) and that d6s are easy to come by.  I have heard, apocryphally, that when Gygax first learned of d20s, he went "This changes everything!"  Having returned briefly to pure-d6 gaming, I can see why.

And this all makes me sad, because I have fond memories of Traveller, but man, actually running games on 2d6 kinda sucks.  It's a pity that Traveller d20 was such a poor port (tangentially, I'm shocked that T20 has such good reviews on DriveThru; maybe I should write a dissenting one).

I feel like converting Traveller to just use d20s (and not also importing the rest of the "d20 System"s baggae like T20 did) wouldn't be hard, exactly - just a slog.